![]() Absence of either of these symbols imply equality as the test. The entry in the JVMVersion key are expected to be a floating point number, optionally prefixed by either a greater-than or less-than symbol. Packages that don't specify a JVM version are assumed to be OK. RBFClassifier is the equivalent of RBFRegressor for classification problems. Checks the supplied package against the JVM version running Weka. It is possible to use conjugate gradient descent rather than BFGS updates, which is faster for cases with many parameters, and normalized basis functions instead of unnormalized ones. RBFRegressor implements Gaussian radial basis function networks for regression, trained in a fully supervised manner using WEKA's Optimization class by minimizing squared error with the BFGS method. If the class is nominal it uses the given number of clusters per class. Symmetric multivariate Gaussians are fit to the data from each cluster. It uses the k-means clustering algorithm to provide the basis functions and learns either a logistic regression (discrete class problems) or linear regression (numeric class problems) on top of that. RBFNetwork implements a normalized Gaussian radial basis function network. There are currently three classes in this package: RBFNetwork, which trains the hidden layer in an unsupervised manner, and RBFRegressor and RBFClassifier, which are fully supervised. ![]() Waikato Environment for Knowledge Analysis (WEKA) RBFNetwork: Classes that implement radial basis function networks.
0 Comments
Leave a Reply. |